Voxta v1.3.1: AI Character Generator, Live Previews & Massive Module Expansion

Hello everyone,

Voxta v1.3.1 has arrived! This release is packed with major additions, focusing on a brand-new AI Character Generator, real-time visual feedback for image generation, and a huge expansion of supported AI providers. We’ve also introduced EchoTTS for high-quality voice cloning and the new Imagine Flow to bring your roleplay to life visually.

✨ New Features & Integrations

AI Character Generator

You can now generate characters from scratch directly within Voxta. This feature includes vision analysis for avatar uploads, making character creation faster and more intuitive.

Imagine Flow

Turn on the "Imagine" option to automatically generate an image after every message, creating a continuous visual narrative for your chat.

New AI Modules

We have significantly expanded our integrations. You can now use:

Google: Gemini (text, vision, image generation).

Anthropic: Claude (text, vision, image generation).

xAI: Grok (text, vision, image generation).

MistralAI: Full suite support.

Tavily: A dedicated search module for better web retrieval.

EchoTTS

A new diffusers-based voice generation module with excellent voice cloning capabilities. Big thanks to @MrDragonFox and @starmoose for this addition!

Chatterbox Turbo

Updated integration for faster performance.

🎨 Image Generation Upgrades

Live Previews

Watch your images come to life in real-time! We've added live latent previews and progress reporting directly in the Chat View and Local Diffusers module.

Local Diffusers Enhancements

Added support for HiRes Fix upscaling, refiners, a dedicated device selector, and improved VRAM/import stability.

Chat Gallery

A new gallery view in the chat explore section lets you easily browse all your generated images.

OpenRouter Image Support

Full support for image models via OpenRouter.

⚙️ UI and Workflow Improvements

Scenario Editor Overhaul

A major redesign splitting Events and Buttons. Includes a new StageView with a position editor, plus responsive, distortable, and scalable button systems.

Playground Polish

ImageGen Playgrounds now persist their state and feature a new interactive PromptTagsEditor. We've also added "Surprise Me" buttons for random generation.

Document Editor Redesign

Includes a fresh look, "Canvas Runner" polish, SASS/SCSS language support, and context menus for document management.

Audio & Vision Menus

Split Input/Output selection into Service vs. Device for clarity. Added a "Character Generator" source to the Vision Playground.

ActionScript Editor

Expanded modal editor with better navigation (Ctrl+Click imports) and improved code snippet UI.

Real-Time Installation Feedback

Python pip operations now show a real-time progress bar in the UI.

🛠️ Key Fixes & Technical Enhancements

Installer Resilience

Better handling of installer failures, including interactive retries and shell fallbacks for locked data folders.

Local Diffusers Fixes

Resolved scheduler crashes, ungraceful IPC shutdowns, and errors affecting some Illustrious models.

Scenarios

New scenarios now automatically inherit the "Voxta UI" parent.

Form Fields

Added support for Array-of-Objects form fields (utilized by Lovense actions).

Thank you for using Voxta. This update opens up many new ways to configure and enjoy your AI companion - have fun exploring the new modules!